From: Intelligent systems for sitting posture monitoring and anomaly detection: an overview
Techniques | Advantages | Limitations | Refs. |
---|---|---|---|
Supervised | High success rates Interpretation of results | Limited number of anomalous samples: Unbalanced data base Health specialist presence required for data labeling Inability to characterize all anomalies | |
Semi- Supervised | Normal samples available Ability to detect unknown anomalies | Since the postural pattern may be composed of different normal states, the normal boundary is wide Expert knowledge required in case of wanting to label different normal states | |
Unsupervised | No data labeling required Detection of unknown anomalies Applicable to large data sets | Increased tendency to false positives Lack of interpretation Normal data are grouped in clusters assumption | [92, 117,118,119, 122, 123, 125,126,127, 129,130,131,132,133,134,135,136,137] |
Unsupervised | No data labeling required Detection of unknown anomalies Applicable to large data sets | Increased tendency to false positives Lack of interpretation Normal data are grouped in clusters assumption | [92, 117,118,119,120, 122,123,124,125,126,127, 129,130,131,132,133,134,135,136,137,138] |